works on

From the 1 of 42 linked papers with an AI index.

activity
20242026
collaborators

42 papers

cs.CV2026

Is Visual Prompting All You Need? Studying VLM Spatial Reasoning under Progressive Visual Scaffolds

Lars Benedikt Kaesberg, Tianyu Yang, Florian Valentin Wunderlich +4

Vision-language models (VLMs) have advanced rapidly in multimodal reasoning, yet recent work shows that their failures often reflect an interaction between visual grounding and dow…

cs.CL2026

Can Agents Deceive? Evaluating Reasoning and Deception in ParliamentBench using a Social Deduction Game

Niklas Bauer, Lars Benedikt Kaesberg, Akiko Aizawa +3

The paper introduces ParliamentBench, an open-source benchmark based on the Secret Hitler game, to evaluate large language model agents on deception, persuasion, and reasoning unde…

cs.MA2026

Misinformation Propagation in Benign Multi-Agent Systems

Jonas Becker, Jan Philip Wahle, Terry Ruas +1

Multi-agent systems, in which multiple large language model agents solve problems through turn-based interaction, are increasingly deployed in high-stakes settings such as medical…

cs.AI2026

Beyond a Single Direction: Chain-of-Thought Disrupts Simple Steering of Refusal

Kia-Jüng Yang, Dominik Meier, Jiachen Zhao +2

Large reasoning models (LRMs) generate chain-of-thought (CoT) traces before producing final outputs, introducing a dynamic internal state that may complicate control mechanisms suc…

cs.IR2026

Aspect-Aware Content-Based Recommendations for Mathematical Research Papers

Ankit Satpute, André Greiner-Petter, Noah Gießing +4

Content-based research paper recommendation (CbRPR) has seen advances in computer science and biomedicine, but remains unexplored for mathematics, where paper relatedness is more c…

cs.AI2026

Multi-Agent Reasoning Improves Compute Efficiency: Pareto-Optimal Test-Time Scaling

Florian Valentin Wunderlich, Lars Benedikt Kaesberg, Jan Philip Wahle +2

Advances in inference methods have enabled language models to improve their predictions without additional training. These methods often prioritize raw performance over cost-effect…